J. D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting additive models. J. Multivariate Analysis, 73:166--179, 1998. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Fitting a Bivariate Additive Model by Local Polynomial.. - Jean Opsomer Iowa (1997) (7 citations) Self-citation (Opsomer) (Correct)
No context found.
Opsomer, J.-D. and D. Ruppert (1995). A fully automated bandwidth selection method for fitting a bivariate additive model. Preprint 95--32, Department of Statistics, Iowa State University.
A Root-N Consistent Backfitting Estimator for Semiparametric .. - Opsomer, Ruppert (1999) Self-citation (Opsomer Ruppert) (Correct)
No context found.
J.-D. OPSOMER and D. RUPPERT (1998). A fully automated bandwidth selection method for fitting additive models. The Journal of the American Statistical Association 93, 605--620.
Estimating an Unknown Function by Local Linear Regression when.. - Opsomer (1996) (1 citation) Self-citation (Opsomer) (Correct)
....by spectral analysis. Similarly, global bandwidths (the kind considered here) work much better in practice on observations that are reasonably evenly distributed. For an example of some of the problems associated with gaps in the observations in nonparametric regression, see Opsomer and Ruppert [19]. Note also that (AS.IIA) is still more general than requiring that the observations be on a fixed, equally spaced grid, the usual assumption for the study of nonparametric regression with correlated errors. The AMASE can be used to develop plug in bandwidth estimators, by following the following ....
....definite, as long as H ( 0 for all (Brockwell and David [6] p.120) This condition is guaranteed to hold for kernel regression. This topic will be explored in another article. 4 Estimation of the Variance We will use a generalization of the variance estimator of Opsomer and Ruppert [19]. A simple nonparametric estimate of oe 2 , oe 2 = 1 n n X i=1 (Y i Gamma m(X i ) 2 ; can easily be constructed, where m is estimated by local linear regression and, for model A, backfitting. Let = diagf k ; k = 1; Dg represent the bandwidth for this estimation. If assumptions ....
[Article contains additional citation context not shown here]
J.-D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting a bivariate additive model by local polynomial regression. Preprint 95-- 32, Department of Statistics, Iowa State University, 13 September 1995. Being revised for The Journal of the American Statistical Association.
Asymptotic Properties of Backfitting Estimators - Jean D. Opsomer (1998) (1 citation) Self-citation (Opsomer) (Correct)
....the covariates which guarantee that concurvity does not occur. In principle, the results from Lemma 2.1 could be used to derive conditions on the joint distribution of the covariates for higher dimensional models. Extensive simulation experiments, reported in Opsomer [13] and Opsomer and Ruppert [15], have shown that in practice, a model fitted with local polynomial regression almost always converges to a good (albeit potentially non unique) solution. Concurvity and even lack of convergence can occur, however, when the number of covariates is high and they exhibit strong correlation among ....
....the backfitting estimator. This comes at the cost of a much more complex and computation intensive fitting method, however. Note that the independence assumption is not sufficient to derive the approximations in Corollary 3.2 directly from the normal equations. Theorem 5 of Opsomer and Ruppert [15] derived expressions that do not use Theorem 3.1 and therefore resulted in spurious O p (1= p nh d ) terms in the asymptotic bias. Even though this term was of the same order as the leading bias and variance terms, it was ignored in the asymptotically optimal bandwidth expressions. By ....
[Article contains additional citation context not shown here]
J.-D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting additive models by local polynomial regression. Journal of the American Statistical Association, 93:605--619, 1998.
Nonparametric Regression with Correlated Errors - Opsomer, Wang, Yang (1999) (1 citation) Self-citation (Opsomer) (Correct)
....not hold for the additive model. For the case D = 1, Opsomer [51] develops a plug in bandwidth selection method that generalizes the Direct Plug In (DPI) bandwidth selection proposed by Ruppert et al. 58] for the independent error case and extended to the additive model by Opsomer and Ruppert [54]. The method described here is therefore referred to as Correlation DPI (CDPI) and was used as the correlation adjusted bandwidth selection method in Figure 1. The estimation of the 22 (k; l) in CDPI is analogous to that in DPI. Opsomer [51] shows that the term IC n is approximately equal to ....
J.-D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting additive models by local polynomial regression. Journal of the American Statistical Association, 93:605--619, 1998.
Estimating a Function by Local Linear Regression when the Errors.. - Opsomer (1995) Self-citation (Opsomer) (Correct)
No context found.
J.-D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting a bivariate additive model by local polynomial regression. Preprint 95--32, Department of Statistics, Iowa State University, 13 September 1995. Revised for The Journal of the American Statistical Association.
Asymptotic Properties of Backfitting Estimators - Jean D. Opsomer (1998) (1 citation) Self-citation (Opsomer) (Correct)
....does not occur. In principle, the results from Lemma 2.1 could be used to derive conditions on the joint distribution of the covariates for higher dimensional models for guaranteeing that concurvity does not occur. Extensive simulation experiments, reported in Opsomer [14] and Opsomer and Ruppert [16], have shown that in practice, a model fitted with local polynomial regression converges to a good (albeit potentially non unique) solution. Concurvity and even lack of convergence can occur, however, when the number of covariates is high and they exhibit strong correlation among them. In this ....
....sense, backfitting does not share the oracle property of the marginal integration method discussed by Fan et al. 4] Both the analytical tractability and the oracle property come at the cost of a much more complex and computation intensive fitting method, however. Theorem 5 of Opsomer and Ruppert [16] derived expressions that do not use Theorem 3.1 and therefore resulted in spurious O p (1= p nh d ) terms in the asymptotic bias. Even though this term was of the same order as the leading bias and variance terms, it was ignored in the asymptotically optimal bandwidth expressions. By ....
[Article contains additional citation context not shown here]
J.-D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting additive models by local polynomial regression. Journal of the American Statistical Association, 93:605--619, 1998.
Regularization Methods for Additive Models - Marta Avalos Yves (Correct)
No context found.
J. D. Opsomer and D. Ruppert. A fully automated bandwidth selection method for fitting additive models. J. Multivariate Analysis, 73:166--179, 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC